Research And Realization For 3D Scene Roaming Of Point Cloud Model Based On Skeleton Line Extraction Algorithm

Three-dimensional scene roaming is the most effective overview of the general situation of the three-dimensional scene.With the continuous development of 3D scanning technology,the application of scattered cloud model is also expanding.It is an important research direction of 3D scene visualization that how to get the structure information of 3D point cloud model quickly and accurately,and to achieve accurate and effective scene roaming.A threedimensional scene roaming method based on one-dimensional skeleton line information of point cloud model is presented in this paper.The first step of the method is to obtain the skeleton line information of the point cloud model,and then abstract the information as undirected graph,then use the graph search algorithm to obtain the roaming path,and finally through the view transformation to complete the point cloud model scene roaming.The acquisition of the skeleton line information of the cloud model is the core of the scene roaming.After analyzing the basic idea and merits of the existing skeleton line extraction algorithm,a simple and effective skeleton line extraction algorithm of point cloud model based on neighborhood search is presented in this paper.The algorithm uses the feature of the R-neighborhood point set and the K-neighborhood point set to perform iterative updating of multiple central points' position,and then uses the special method to select and connect the skeleton line points to complete the whole skeleton line extraction.The algorithm has a number of tunable parameters,has a strong applicability,and high efficiency.We get the weighted undirected multi-graph by abstracting the skeleton line information,accordingly,the global and fixed-point roaming path planning problem of point cloud model is solved by using the improved Tremaux search method based on the depth-first search and Dijkstra algorithm.Finally,the scene roaming of the point cloud model is completed by giving the viewfinder parameters and performing the view transformation according to the skeletal path to update the view and sight direction.The experimental results show that the method of scene cloud roaming proposed in this thesis is highly automated and has high path accuracy.